- il y a 1 jour
The European Institute of Planted Forest (IEFC) is hosting a series of webinars to introduce the project’s Innovation Actions to the public.
Fourth session: “Next generation tools for predicting, simulating & managing wildfire risks”
Fourth session: “Next generation tools for predicting, simulating & managing wildfire risks”
Catégorie
📚
ÉducationTranscription
00:00:00thank you very much so yeah i'm gonna i'm gonna share my screen
00:00:05can you see my screen right yes it's perfect you can start please okay um so yeah adrian
00:00:18did a great job explaining the the umbrella system uh the platform we develop in fibrous
00:00:26so the objective of that platform is to integrate all the innovation actions or all the technological
00:00:33innovation actions we have in fire res and and and to to better show how they work in in the in the
00:00:40lab so that's the primary goal of the umbrella system and also uh do a kind of pilot pilot well
00:00:48in in fact it's not a pilot it's a an operational tool that can be used for operational decision
00:00:54making and and then with that tool we wanted to integrate all the info our fire analysts our
00:01:03advanced need to support the decision making uh by integrating copernicus the fire risk fire
00:01:11danger indexes the volatile compounds the weather data and so on so in this sense they have all
00:01:18the data needed to support the decision making um right now during my presentation i'm gonna show
00:01:27uh the main innovations we have regarding the fire danger and fire risk and how we are model
00:01:35the fire risk dynamically so the idea at the beginning of this project was to go beyond a fire danger
00:01:44this is the kind of input output or the kind of data we have in copernicus and going beyond that
00:01:50and estimate the uh dynamic risk and consider the the assets at risk the population the buildings
00:01:57threatened the buildings destroyed and so on and uh as adrian mentioned the idea is to uh model the
00:02:05real-time fire behavior uh to quantify the the asset impact so uh it's not just a static risk assessment
00:02:14uh we are simulating the the five behavior dynamically based on real-time weather conditions
00:02:22so this is just to mention that we are in tecno silva we are a team of two two hundred plus people
00:02:29working on real-time welfare modeling and risk analytics working with different agencies in europe
00:02:35but also in the us with cal fire and british columbia among many other customers in the netherlands ireland
00:02:42italy and and so on and the idea is that to transfer all the knowledge from fire agencies to this
00:02:50project and and the opposite uh get and leverage all this work done in doing the fire risk project to
00:02:59provide the solutions to the to the fire agencies so basically this is the work package five so this is
00:03:06the fire arrest umbrella we have um around 10 work packages and i'm going to explain some innovation
00:03:14action of developing the work package five that is about uh technological solution decisions support
00:03:21systems to support stakeholders uh dealing with extreme warfare events in europe more or less that's the
00:03:27the idea uh so yeah this is just a summary uh at the beginning of the project we were working with a
00:03:36fire danger products from ephys from copernicus uh the the platform is great the website is great and
00:03:43the data they provide is very useful uh we can track the the hotspots we can see the different fire
00:03:51danger indexes such as fire weather index and also other drought indexes but it's true that we have
00:03:59uh some limitations so we don't have we don't consider the impacts so we don't have risk so we only have
00:04:07fire danger so we don't quantify the the risk to specific communities or or assets
00:04:15these kind of metrics are where they're only focused uh and it simplifies the complex interaction
00:04:21between topography fuel types and weather so basically we only consider the weather data
00:04:27and it's more beyond weather we need to consider the fuels the topography and how those complex
00:04:33interactions impact the the assets or the or the assets at risk such as the population of the of the
00:04:40buildings then we also have lack of temporal variability uh this kind of outputs are produced daily and we
00:04:48cannot capture changing fire conditions in real time and then uh the the validation is limited in some
00:04:56areas especially at local scales uh and but the boy the the best point is that we cannot quantify the
00:05:03the risk and we miss some important factors such as topography and fuel ties so this is the the original
00:05:11starting point and and from that uh we wanted to improve it so fire danger indexes are great uh but we
00:05:20wanted to make them more uh operational to inform decision making so uh during this project and also
00:05:29in collaboration with fire agencies such as cal fire or the or other agencies in europe we developed the
00:05:36both indexes to uh support the initial attack and extended attack so these are the initial attack
00:05:43assessment index and extended attack assessment index so these indexes include weather fuels topography
00:05:52based on simulations and also atmosphere stability uh to consider the pyro convection so the idea is to
00:06:00better support decision making in the different phases the initial attack and the extended attack
00:06:05and also adjust them to uh consider extreme welfare event uh fire behavior uh in this case to consider the
00:06:13the the pyro convection so in that sense we develop uh different metrics to estimate the probability of
00:06:20pyro convection that fit this kind of modeling so i will show you more details later uh obviously these
00:06:27metrics are not risk metrics since they we are not considering the values at risk but uh
00:06:34we are modeling the the indexes every three hours and we validated the indexes with more than 30 000 wildfires
00:06:46in the us but also in in europe and also uh we also wanted to evaluate the fire risk dynamically to
00:06:56consider these complex interactions between topography fuel ties weather and atmospheric stability so that is
00:07:04important because when we consider the values at risk and this reality real time weather conditions we get
00:07:11the dynamic risk assessment four days in advance hourly and all this framework has been validated with
00:07:19more than 5 000 wildfires in western u.s with 15 minute observation i will show you a kind of example later
00:07:27uh and the kind of data we are using to simulate uh and the kind of data we are using to simulate because
00:07:31it's uh unprecedented unprecedented um data to to validate and calibrate the the the fire spread models
00:07:40so this is the umbrella system uh all the all what what what we integrated and what adrian jimenez show
00:07:49earlier and i'm gonna show you how we adjust the fire spread models and the fire danger indexes to capture the extreme wildfire events
00:08:03so basically we uh have uh two things here um as i mentioned we wanna capture atmosphere stability uh to uh adjust the rate of spread
00:08:15the rate of spread according to the to the forecast to do that we have developed and we have analyzed
00:08:22relationships between atmosphere stability indexes such as the lifted index or the convective flag developed by
00:08:30tecno silva um to the to the rate of spread and with that we develop uh some kind of modeling to adjust the
00:08:40the rate of the rate of spread according to the atmosphere stability and better estimate or at least
00:08:47more realistic the expected fire behavior so uh i will start with the fire danger indexes and then i will go
00:08:55through the the the fire risk metrics so as i mentioned we developed two kind of indexes two indexes
00:09:03for analyzing the initial attack and the extended attack uh for the initial attack is is to capture
00:09:12the the probability of success when uh when we have an ignition so every time we have an ignition we
00:09:20can simulate a fire automatically and we are evaluating that fire from one to five uh analyzing the the
00:09:30suppression difficulty in that initial attack so here we validated the index with more than 30 000 wildfires
00:09:38in in in western us and and europe and here you can see how the how the index works in these two two
00:09:45charts so in the first chart we can see how the probability of initial attack success uh is uh depends on
00:09:57the index and we can see that in the highest values of the index we have a lower probability of initial
00:10:04attack success and then when the initial attack fails we can see that the index is gonna tell us
00:10:10uh about the final fire size and about the fire potential index uh so larger fires when the index is
00:10:18is high so this is very important because uh with an automatic simulations for any incident we have
00:10:25we have a prioritization so we have a one to five uh index in all the fires and we can prioritize very very
00:10:32easily our dispatch but also uh we can prioritize the the incidence by uh by by by level so that's regarding
00:10:45the initial attack assessment index but what happens when we have an extended attack when we have failed
00:10:51in the initial attack and we have uh five analysts working on the on the fire so to do to work on that we
00:11:00created a matching learning a model uh trained with real fire activity uh so it's uh that we produce
00:11:08every three hours so here we have several variables standard variables are just field complexity
00:11:16or upper pressure deficiencies the terrain difficulty drought indexes but we also have a convection
00:11:23potential um with the convective flag and atmospheric instability with the lifted index
00:11:30to capture uh atmosphere stability and the good point is that we are calibrating or we calibrated the
00:11:38model using real time fire real fire activity from satellite hotspots in both the u.s and europe here is just an
00:11:49example for one fire this is the mckinnon fire in in the u.s and here you can see the the index
00:11:56progression throughout time uh the points uh represent the the real fire activity the hotspots from satellites
00:12:05and here we can see the fire activity and how the fire could be controlled uh when the index went down
00:12:15so and here we can see all the some of the sub components of the index we can evaluate the the
00:12:20drought the atmosphere stability uh the vapor pressure deficit or the fuel moisture wind speed and and so
00:12:28so on and the good point is that we can now capture the convection so here is just an example for the
00:12:36mckinnon fire in the first three days we have a potential activity and we can see how the index
00:12:44uh identify these kind of conditions when the index is very very low such as in in this case we may expect
00:12:52um convective uh five behavior and that's what we observe uh from the cameras and in the in the field
00:12:59and this also feeds the the index and make it uh more higher more more dangerous
00:13:09so that's regarding the five danger metrics uh just very briefly i will go through the
00:13:15fire risk metrics as adrian mentioned uh we can run a simulation every time we have an ignition
00:13:22uh we can run this kind of of of fire simulation um the good point is that in the u.s we have this
00:13:33kind of of service so those polygons you see on my screen uh is the the the real fire progression
00:13:40every 15 minutes we have that data in in real time and we have collected more than 5 000 wildfires
00:13:48uh and and then we can validate the the model performance so two years ago we published that
00:13:56paper uh about the performance of operational fire spread models in in california we have more fires now
00:14:03and and we are uh we can provide um error bias and uncertainty metrics um and also improve and better
00:14:13calibrate the the fire spread models and this is what we have in in fire arrest although we have
00:14:20enhanced this kind of modeling to capture the extreme wildfire events based on atmosphere stability
00:14:28this is uh the oldest fires we analyze in in in in california in this paper but i mentioned we have
00:14:37data right now in in the whole western u.s we have more than 5 000 fires and then we have the the risk
00:14:43metrics this is uh also adrian jimenez showed that um in this case we can calculate the fire risk every
00:14:51three hours in fire res um this is dynamic uh risk assessment we put an ignition point every 500 meters
00:15:00and run independent simulations with our hpcs uh based on real-time weather conditions fields and topography
00:15:09and and and this is great because we can hourly estimate um the fire risk uh across the across the territory and
00:15:25and be threatened the population impacted and and so on basically this is uh
00:15:32uh by using the uh by using the the the higher the hpcs we have to run simulations virtual simulations every
00:15:41500 meters and finally this is the kind of adjustment uh we have done to better estimate the extreme
00:15:49welfare event so fire risk is about extreme welfare events and we made the effort to capture that and to
00:15:57uh learn from from this kind of fires so with that fire guard data with that data with those polygons
00:16:05every 15 minutes and also with uh satellite hotspots uh we develop a rate of spread estimates for every
00:16:14single fire and we relate that with the atmosphere stability and all other variables such as wind speed
00:16:21surface uh wind speed fuel moisture and so on but we try to isolate isolate them the effect of the atmosphere
00:16:31stability and we found a very strong correlation between the lifted index and the convective flag with
00:16:38the rate of spread and we can see that the rate of spread can be can double the the rate of spread when
00:16:46the atmosphere is very unstable and this was what we tried to capture uh in the model so we created
00:16:55or we developed this um preliminary model to adjust the the rate of spread as we recognize that it's
00:17:04preliminary and we are in the frontier of the science we need to work more on that uh but it's the first
00:17:11step uh to operationally capture this kind of phenomena and with this kind of adjustment at least we can
00:17:20create like a probabilistic simulations using variation ranges a variation range in the in the adjustment
00:17:28and get this probability simulation to show the the fire potential when we make when we have this kind of
00:17:36of phenomena and finally very very briefly um so just highlight all the uh high resolution weather data we
00:17:46have we are producing for uh feeding the fire spread model this is just an example about the high resolution
00:17:54weather data uh technosilva is is providing uh two kilometer uh high resolution weather data in 3d at different
00:18:04pressure levels about in this case you can see the the wind speed in a in a in a region but we are also
00:18:11providing both dead and and and light field moisture content among other parameters and as i mentioned we
00:18:18have different initiatives to get the the high resolution fuel maps in this case this was this is led by the
00:18:26institute cartographic of catalonia which uh is providing a high resolution fuel maps to fit the the fire spread
00:18:35models in in california although we also have the panareo european system to define these fuel maps and and
00:18:44also support the the management priorities to mitigate fire impacts and also run simulations anywhere in in
00:18:52europe and yeah and that's it uh thank you so much for the invitation and i'm happy to address any questions
00:18:59you may have okay we don't have so much time remaining but the two presentations were quite in line i don't see
00:19:09any question um so here we are we are also using um a canadian model and to to to model to make this kind of
00:19:18modelizations and for for the firemen uh you know to use them in real time um the most important thing
00:19:27was to see how much realistic it was but i see that most of the validation has been done in california so
00:19:34are the firemen happy with um with the resulting in new in catalonia yeah that's a very good question um
00:19:45unfortunately in europe we don't have the kind of data we have in the us um in the us is great because we have
00:19:52the progression of every single fire every 15 minutes so we can we have good estimates we have the fire suppression
00:20:00resources we have everything and we can better calibrate and badly validate our modeling that's the reason why
00:20:08we have started in the in the us but we have also working in in europe uh issue that we have less data
00:20:17but for instance we have leveraged the the satellite hotspots uh from beers from modis and also working
00:20:25with different reports uh for instance from catalonia that the the firefighters uh develop or or write
00:20:34uh very nice reports so with that data we also uh validated the the modeling but it is true uh that
00:20:44we don't have that huge amount of data and high quality data we have in in the us it's true that we have
00:20:52that limitation in in europe but um obviously for instance with the fighters we had in catalonia in the
00:21:01in the in the in our living lab this summer we are testing all the all the modeling there um and we
00:21:07are going to provide um a kind of report our kind of uh deliverable a kind of presentation to show how
00:21:14the the models work um but it is true that we don't have that amount of data that's the reality but also
00:21:21yeah trying to to to to do the best in europe to to to validate the the performance of of the modeling
00:21:29yes another yeah we had the same problem because we don't have a large fires we had only large fire
00:21:34in 2022 but usually we're only very small ones um the second question is with lidar uh we are considering
00:21:42to to be able to make new type of fuels uh where there is some cleaning in the understory
00:21:50to make different spread propagation different propagation rate or intensity when when it is clean
00:21:56because you know we want to test the efficiency of cleaning measure because there is a strong
00:22:02recommendation for landowners to make some cleaning the forest management includes also some removal of the
00:22:09brushes and so on um is it something you are considering you have fine enough uh fuel types or or there is
00:22:19not really any cleaning done in forests it doesn't make sense in catalonian context
00:22:25well i i think if we want to model uh the fire behavior properly i think one of the challenges we have is to uh
00:22:34uh properly model the the fuel types so uh i totally agree with you in that sense that we need to model
00:22:44for instance the dynamism of the grassland areas so this is something we are doing in the u.s
00:22:49we are modeling the the grass fuel loading across the landscape uh dynamically to to reflect the the
00:22:58reality because uh the dynamism is important and the grass fuel loading changes a lot throughout time
00:23:06based on precipitation drought and things like that and that's one thing uh but i agree with you that we
00:23:13need to capture the understory of those fuels it's not the same a timber understory fuel types type compared
00:23:20to a timber leader fuel type and that change is important uh so with lidar uh we can capture that
00:23:29also it depends on the quality of that lidar data when we have enough uh points by a square meter we can
00:23:38capture that and we uh are using that kind of data to uh say or to identify when we have a tu model or a
00:23:48timber leader fuel type uh but yeah this is something we need to incorporate more in the in the in the model
00:23:56because it it is essential it because it's gonna it's gonna change um the surface rate of spread but
00:24:04also the transition from surface to crown fire behavior and that's the the big difference it's not
00:24:09the same to have a surface fire compared to a passive or an active crown fire okay thank you so just to
00:24:18let you know we this is exactly the work that we that has been implemented within in the living lab of
00:24:24aquitaine to have this this new new fuel types with with and without understory so okay yeah thank you great
00:24:33just uh i think we are done we are just on time for for the next speaker um so we have no more adrian to uh to
00:24:43ask great this time it's jordina uh so we will talk about uh air polluants uh exposure during wildfires uh so jordina uh the
00:24:57floor is yours so you can start thank you hello everyone
00:25:06uh do you see my screen my presentation in full screen yes it's perfect you can start okay
00:25:13thank you thank you so hello again my name is jordina i'm a phd student at the diathesis and i will
00:25:21present how we quantify the impacts of exposure to air pollutants from wildfires in the framework of
00:25:28innovation action 5.7 so understanding of smoke exposure concentrations and scenarios is crucial for the
00:25:39development of mitigation strategies to improve improve public and occupational health and safety
00:25:47during wildfires and prescribed burns especially in this climate change scenario where wildfires are
00:25:55becoming more frequent intensive and and extend and so what do you want to know what do we want to know
00:26:03it's if all the smoke that are emitted from wildfires are biomass burning so from 2022 to 2024 we monitored
00:26:13the firefighter exposure during prescribed burns and some wildfires in the catalonia living lab the
00:26:20northeast of spain and the study focused on two main firefighter roles the tortures who ignite the fire
00:26:28the the fires with a dripping torch and the liners the line operator who manages the the fire perimeter
00:26:37we use two types of microethylometer the ag51 and the ma200 to to for the black carbon measurements
00:26:46which here in this in this in this figure we can see the ma200 and this uh this device allowed us to
00:26:57perform a source apportionment model to differentiate biomass from the fossil fuel emissions for the
00:27:06particulate matter we use this other little device which has the name of air beam
00:27:16so first of all i will show how we got this type of data and how we monitor the firefighter exposure
00:27:27first of all we reprogrammed the the sensors to eliminate the need for the the mobile application
00:27:33we tested real-time data collection in prescribed burns and we created the infrastructure for real-time
00:27:39data collection and implemented technosilva access for real-time and historical data as the two previous
00:27:47presentations show and this is the type of data that we that we get from the portable monitors on the
00:27:56right we can see against a study in a prescribed burn case and in a wildfire in the map shows
00:28:04the firefighter track and yeah this is for a single person however in prescribed burns we monitor an
00:28:13average uh from four to eight firefighters and we can see that exposures are different depending on the
00:28:21different tasks performed some are tortures and some others are line operators as we can see in this video
00:28:31here is an example of how we monitor the exposures in a prescribed burn with the portable monitors and on
00:28:40the on the right we can see the the representation the the representation of the smoke that the sensors that
00:28:51we had in galicia in the living life of galicia reached the the sensors the smoke of of the the wildfires
00:29:02in the northern portugal in the last summer of 20 in the last of the last year sorry reached the our sensors in galicia
00:29:09and we can see the how these high concentrations were seen by our monitors
00:29:21here i will show you some brief results from these campaigns these results are also published in the
00:29:31the environmental pollution journal and briefly we can see here the mean pm 2.5 exposure concentrations
00:29:41of some firefighters task overall we can see that torture show with the highest mean pm 2.5 exposure
00:29:49followed by the line operators which showed lower mean exposure concentrations but still notable
00:29:55in the
00:30:00more pop tasks which involve involve the final extinguishing using manual tools that may stir up
00:30:05the particles from the ground were identified as a relevant contributor to pm 2.5 exposures
00:30:13also but this is for the prescribed burns for the wildfires we did not have
00:30:20have enough representation of cases to perform more detailed analysis however stands out the
00:30:27direct attack task regarding the black carbon exposures here we can see some similar results
00:30:40and the mean black carbon exposure concentrations and again we can see that tortures
00:30:46show the highest mean exposure concentration followed by the line operators
00:30:55with the black carbon data we could perform a source apportionment the source apportionment model
00:31:02in which we we we detected that 77 overall of the exposure of tortures was attributed to fossil
00:31:11fuel emissions suggesting that they were likely involving the use of the drip torch exposed them to
00:31:18black carbon from fossil fuel in contrast biomass burning derived black carbon had a greater
00:31:25impact on onliners finally uh here i added a slide regarding our latest work on polysilky aromatic
00:31:35hydrocarbons which are a large group of organic chemicals generated through the incomplete incomplete
00:31:42combustion of organic matter which makes wildfire a major source of this of these compounds they also
00:31:52stand out for the environmental persistence and their significant health uh implications especially
00:32:02the venzoipirine which is considered carcinogenic to humans by the international agency for research
00:32:09on cancer so again here we see in these box plots we see the mean black carbon exposure concentrations and
00:32:17the sum of the uh policy of the polycyclic aromatic hydrocarbons and again tortures face the highest mean exposure
00:32:27concentrations followed again by the liners and and the trackers who had the lowest exposures
00:32:35uh a key funding was that the the strong correlation between black carbon and the and the sum of the
00:32:44polycyclic aromatic hydrocarbons which uh correlates very good indicating that black carbon may could be used
00:32:54as a as a proxy for the phs measurements which uh may be more cost effective and and more reliable
00:33:06and these these other results are published in this chemistry journal so if you want to if you have
00:33:15a curiosity for these results here are more detailed and thank you very much and i wanted also i want
00:33:22also to thanks the the the close collaboration with the bombes de la generalitat de cataluña and the
00:33:30dirección general de defensa de monte de la junta de galicia and thank you very much thank you very
00:33:36much you have been very quick you have saved i know so this is a very technical question about the exposure of
00:33:46staffing fires i don't know if there is any question i don't see any hand up
00:33:53um okay so me i have only one question it's um how did did you select the the polluant to monitors
00:34:05uh do you do you know if there is any other kind of chemicals that would be quite toxic to to to check or
00:34:13or or did you already made the screening and identify the one yeah well pihs uh are uh there are uh
00:34:23the you know the international agency for research on cancer um classified pihs uh some are probable
00:34:32carcinogenic for humans and others are carcinogenic to humans so we wanted to focus to these uh
00:34:39harmful pollutants for for for the human health so that's why we selected these these pollutants okay
00:34:48so yeah yeah you have done this working and of review now okay um from my side no question i don't see
00:35:00any other question in the chat uh so thank you it's very much and we will move to the last speaker
00:35:08it's uh victor uh so victor race will talk about prototyping and testing innovative to
00:35:18price when you win event training certificates so again we we talk about firemen but this time
00:35:26about how to train them so yeah good morning slideshow you can start at the moment we see your email
00:35:36software we don't see your slideshow
00:35:54okay now we see the slideshow it is not in presentation mode right we can switch it
00:36:00the presentation mode
00:36:05you can launch the slideshow
00:36:09yes it's perfect you can start thank you okay
00:36:18so good morning everyone my name is vitor and i'm representing the portuguese national fire service
00:36:25school um it's a pleasure to be here and to present the outcomes of the innovation action 5.8 of the fire
00:36:35rest project this action was developed by our school in b together with the fireware partners the catalan fire and
00:36:45rescue service techno silver and the portuguese national authority for emergency and civil protection
00:36:53each bringing their specialized knowledge and experience in this presentation i will take you through the
00:37:01the journey we made in designing piloting and evaluating a new training program specifically
00:37:09target target at incident commanders and fire analysts
00:37:14two key roles in the management of extreme wildfire events
00:37:21so as already mentioned in previous presentations extreme wildfire events are becoming
00:37:28one of the most critical natural hazards
00:37:30and the fires are more intense and more unpredictable than in the past and to respond to this kind of
00:37:40events you require advanced competences and capabilities decisions must be made relatively under conditions of
00:37:51uncertainty and with high social and political pressure
00:37:56incident commanders are the the center of strategic and tactical decisions on the fire ground
00:38:06and the fire analysts provide the scientific and technical inputs that support those decisions
00:38:14but current training often does not fully prepare them for the complexity of this extreme scenario
00:38:23and this was the starting point for this innovation action
00:38:30our main objective was clear to reinforce the collaboration between incident commanders and fire analysts
00:38:39in the management of extreme wildfire events and to do this we applied an instructional design methodology
00:38:46the ADORA model which means analyze design organize develop and evaluate
00:38:56so this methodology allowed us to move from a careful analysis of needs training needs
00:39:05to the design of learning objectives to the organization of the course structure
00:39:11the development of material the development of material and finally to the evaluation of the course
00:39:18so you can find this work in two deliverables deliverable 1.2 that addresses the training program design
00:39:27and deliverable 5.11 regarding the training program delivery and evaluation
00:39:33so we started by analyzing in detail the training needs and capability gap for incident commanders we found
00:39:46challenges in predicting fire behavior coordinating multiple fire funds handling huge amounts of information
00:39:56and managing stakeholders in high pressure context
00:40:02for the fire analysts the gaps were linked to the limited use of predictive tools
00:40:09to the difficulties in handling real-time meteorological and fire data
00:40:15and a lack of clarity in how their role integrates into operational command structure
00:40:23and then we looked at the IC and the FA working together and in that joint work we identified weak points
00:40:38in situational awareness in the communication and in the adaptive planning so these gaps became the foundation for the training design
00:40:53the program that the program that we have designed assisted of 14 lessons delivered through a combination of methods
00:41:04lectures case studies with group discussions demonstrations and immersive simulations the content covers both role specific skills
00:41:18for example how fire analysts use predictive tools or how incident commanders structure their decisions
00:41:29and also the shared skills between the two roles such as communication, situational awareness and adaptive planning
00:41:39the assessment tools were built into the program from the start using observation grids
00:41:48that were tailored to each of these roles to ensure that competencies could be measured during the exercise
00:42:00this design was tested in practice during the pilot course
00:42:05the pilot course that was held in the portuguese living lab in september october 2024
00:42:15at the enb forest fire specialized training center located in lausanne
00:42:23it was a five-day training course with 34 hours
00:42:28bringing together pairs of incident commanders and fire analysts
00:42:33from italy spain portugal sweden and jordan and jordan since this was not an introductory course
00:42:43but a specialized program the trainees were already experienced professionals in this role
00:42:51and were certified if this certification exists in their own countries
00:42:58the trainers the trainers came from the fire west project partners so from our school from the portuguese national authority for emergency and civil protection
00:43:12from the catalan fire and rescue service from technosilva
00:43:15technosilva and we also had the support of a meteorologist from the nipv in the netherlands
00:43:26this pilot course allow us not only to validate the training design
00:43:31but also to gather reach feedback from both trainees and train
00:43:39one of one of the most innovative aspects of this course was the integration of several technological tools
00:43:46so we used the fire sim for simulating the fire behavior
00:43:52we based the training on the version of the tool that was available at that time
00:43:57we also used a tool called web monitorization that was developed by the portuguese civil protection for operational monitoring and decision support
00:44:10we used virtual reality commercial software for creating immersive wildfire scenarios and also some media simulation videos
00:44:26and we used videos and photos from real wildfire events to connect the training with the real world pages
00:44:35so all these tools brought realism interactivity and allowed the trainee to make decisions as if they were in an actual emergency
00:44:52the course content was enriched by several case studies so we asked the trainees to bring their own experience to the course
00:45:05so case studies from italy spain portugal sweden and germany
00:45:10and also to show the diversity of wildfire context across europe the course
00:45:19also reinforces the operational responsibilities of incident commanders and fire analysts
00:45:27introduced structured approaches to decision making and to information management
00:45:35and trained the participants to apply decision support tools such as the the meteorology tools the polygons
00:45:46the web monitorization fire sim and also the virtual reality simulation software so this mix of theory lessons learned and tool practice
00:46:04and made in our opinion the course both comprehensive and practical
00:46:13at the core of the course were three large scale simulations these simulations were based
00:46:21on some devastating log fires of 2022 and 2023 that happened in portugal
00:46:31and these exercises recreated dynamic fire spreads scenarios and the pairs of incident commanders and fire analysts
00:46:45had to process during the exercises the incoming data they had to use predictive tools adapt strategies and their
00:46:56tactical tactical plans and communicate effectively under stress after each simulation we held structured debriefing
00:47:10to analyze decisions to highlight also the good practices and to reflect on what the training could improve for the next exercises
00:47:25and the trainees reported that these hands-on simulation exercises were the most valuable part of the course
00:47:37to measure the learning outcomes we used the role specific evaluation grids so for the incident commanders
00:47:45we assessed uh we assessed uh we assessed uh information gathering the planning the communication and command
00:47:52decisions the fire analysts were assessed uh one day use of meteorological data fire behavior knowledge
00:48:03and the tactical planning and we also assessed the collaboration between the two
00:48:10which which which which was uh of course central to this program uh the results was clear uh all trainees
00:48:19reached the required the required competences by the end of the pilot
00:48:26mainly because they were all very experienced in this world
00:48:32the trainee feedback about the course was very encouraging um 75 percent reported uh being satisfied
00:48:47and 25 percent were very satisfied with the with the course they found the course highly relevant to their work
00:48:57and particularly uh appreciated the use of the polygons methodology
00:49:05um the meteorological data and also the simulation exercise but at the same time they suggested adjustment for instance
00:49:17uh we should dedicate less time to the tool fred monitorization and to the lesson uh operations management
00:49:30and more time to the fire analyst role and also to practice with the decision support
00:49:38behind the satisfaction surveys uh we also collected uh some qualitative feedback and the strength uh identified uh included
00:49:55the practical simulations the integration of innovative tools
00:49:59and the collaboration between the ICs and the FAs the challenges included the diverse background of the participant
00:50:12the fact that some tools uh that we use uh are only available in portugal
00:50:19and the need for better translation and the preparatory materials that were used during the course
00:50:30so this the recommendations were clear so we should increase the hands-on practice and adapt the course for the European context
00:50:44with this feedback we had a clear pathway for the refinement of the course so future courses will expand time
00:50:54for polygon for polygon for the meteorology for the meteorology and for the information management we will tendance theoretical lessons on IC and FAO
00:51:09um we will include a module uh regarding leadership and communication and uh also in future editions we will have to improve logistical and uh some administrative aspects
00:51:29uh including the materials that i already mentioned so this refinements uh will make the course more efficient and uh
00:51:38uh accessible uh accessible uh across europe
00:51:43the key takeaways from this innovation action um are basically the following
00:51:48so we have uh innovative tools
00:51:51and simulation that enhance the preparedness of the incident commanders and fire analysts
00:52:02the heads-on ends-on scenario based training is the most effective learning approach
00:52:08for this kind of approach for this kind of course the collaboration between ICs and FAs is critical for managing extreme
00:52:17warfare events and finally replication at european level requires some careful adaptation to different contexts
00:52:26projects uh and different systems uh and different systems
00:52:30the conclusion um well in our opinion the pilot course uh was both effective and well received by the the trainee
00:52:42and uh the ultimate ambition regarding uh this innovative uh action is to publish a european certification pathway
00:52:51for the ICs and the FAO training uh and this of course will provide recognition of skills and um
00:53:02friends the europe's resilience to this kind of uh warfare event so thank you
00:53:10thank you thank you very much i can confirm that the firemen who attended one of your events they were very happy to be there
00:53:23from the french living lab so let's see if there is a couple of questions or not
00:53:28um yes so there is a question from gabriella gabrielle um i think you can turn your microphone on to make the question
00:53:40as we have quite a bit of time um so if you want to make your question i can read it unless
00:53:49yeah really okay no okay i will read the questions then
00:53:52it's about fire steam um it's about fire steam i think it was a question for the previous speaker
00:54:00because it's about yes i think so simulation stochastic or the the determined determinist um i don't know
00:54:11if someone one of the adrian would like to answer and then we will come back for question to rito
00:54:18um so madrian i think you're right knocking that answer but uh here's miguel mendes also from tecna
00:54:27silva okay um i can provide an answer i think uh so regarding fire sim software so we uh use and
00:54:36implement it also in the integrated integrative software system that we presented um fire sim with
00:54:43the two approaches so um stochastic uh simulation uh operational as well because it's quite fast it
00:54:52provides results takes a bit longer than the other the standard approach but um it simulates uh in a
00:54:59couple of minutes two to three minutes uh stochastically also the simulations and um also then the
00:55:07the other one the more standard operational uh model um that is not stochastic so we have both
00:55:15options and both options can can be used uh when triggering a simulation maybe the user can select the
00:55:22the the model that they want to use and they trigger the simulation to see the corresponding results
00:55:29okay okay so we have both um there is a question from conception you can ask it uh directly because
00:55:39so we have time to take to have you turning your microphone on and making the question for vito
00:55:46thank you for the presentation so you talk about you did some scenario with the media uh i think like
00:55:54if it was a media broadcast uh i'm wondering uh what did you do how did you explore it and let me just put
00:56:02the plug because my computer is going down wait yes uh we have used uh also the simulation software
00:56:12to produce uh some media clips and um the media uh was a way to inject uh more information
00:56:22into the into the into the command port um to see how the incident commander or even the fire analyst
00:56:31could react to the to the information that that was coming from the the fire ground so it was a way
00:56:40of keeping the simulation running and also adding some information using that kind of source
00:56:52uh can i ask another question yes yes please go on okay but did so the incident commander had to
00:57:03then have some ideas how to interact with the media or that wasn't the purpose
00:57:11no that was not the purpose to have a kind of media training uh it was just used
00:57:18uh as an inject of information uh it was not continuously during the simulation it was just a moment in each
00:57:29exercise where they they were aware of a fire development or a city emergency situation that was what got
00:57:40broadcasted by by media
00:57:48okay me i have a couple of questions then there is no more questions um one is about the virtual reality
00:57:56is it something that is really real virtual reality i mean you just go on the field and you see flames or
00:58:03fire arriving or fire arriving or is it just kind of video where people see themselves in the video or how do you
00:58:11use this kind of tools and well this software that we use creates a 3d scenario and we can
00:58:22uh move into the scenario and we can see where the fire is progressing and so we can have a quite immersive experience
00:58:36regarding the fire progression we selected some scenarios that are quite similar to the
00:58:44uh to the places where the fires actually uh so the trainees had a sense of being uh in place uh through the that software
00:59:03okay and so this is something that can be shared or is it something you're just talking commercial softwares
00:59:08it is commercial um from our side what we have done is developing the scenario because the scenarios are not
00:59:20pre-built into the software so we had to to create uh the scenarios adapted to the exercises that we have
00:59:30running okay i have another question about this training how much um real-time modeling is used
00:59:38in in portugal for example in france with large fire in in aquitaine in 2022 and the fireman adds the
00:59:46modeling and in 2025 in odd they are not using any models at all so just you know if it is something
00:59:55that is used everywhere in portugal or in spain or is it homogeneous all over the country or is it
01:00:01depending on on the on the regions or the area so i have lost you a little
01:00:19i'm sorry i didn't understand uh your question yes my question is um in the training you propose
01:00:25to the forest to the forest to the firefighters to to have uh to use real-time modeling so that they can
01:00:32see how the fire will evolve and so on and how the the firefighter can can uh behave and so i was
01:00:43saying that in france we have in the in the demo sites that people are using a lot of modeling but in
01:00:49in the large fire we had in pyrenees they did not use it because it is very heterogeneous it really
01:00:55depends on the on the areas not all the firefighters over france are trained to that and i just wanted
01:01:02to know how is it in portugal or in spain is it something that is used everywhere or is it just a few
01:01:10of the forest fire firefighters uh we are aware of the modeling tool in real time just just to understand
01:01:18where we are sure so in training we are using the modeling software in our training center so we
01:01:27train the portuguese fire officers to use the tool and to use the information that is provided
01:01:37by the modeling so they can include that in the decision making process
01:01:44in real fires in real fires i suppose that some use it others don't so the reality is not
01:01:58totally homogeneous at least in portugal
01:02:04okay and in spain you have some spanish colleagues is it widely used in real fires or it is still
01:02:13considered as research stuff or for example in catalonia do they use it already
01:02:22i think technosilva can better answer that question yeah yeah yeah thanks for the question i think it
01:02:30depends on the on the region some regions for instance have implemented our software for instance and
01:02:38andalusia castilla and leon and others and they are using it uh for operational decision making
01:02:46in other regions uh they don't have this kind of solution so it depends on the on the region
01:02:52in the region okay so yes everything is quite heterogeneous there is still a lot of work then to to spread
01:03:00all these tools and innovations all over europe i think this could be quite a nice conclusion uh we
01:03:06have done significant steps within fire res and there is now a lot of effort to do to to spread that
01:03:13uh you know our countries um so i see there is no more question i just would like to thank all of you
01:03:21and i think it's time uh for us to close this webinar it was quite long but it was quite interesting
01:03:28we have seen a lot of interesting innovation um it might be the final webinar uh because then there
01:03:36will be a final event in november so we will see if we are able to organize a webinar in between or not
01:03:43in any case i would like to thank all the speakers and um all the audience and um we will try to put
01:03:51this webinar online uh in the coming weeks thank you very much and have a nice day bye bye
01:04:00thank you bye bye bye thank you everyone goodbye thank you so much
01:04:08have a good day
Recommandations
1:11
1:06:59
1:24:49
1:59:16
1:31:51
1:45:13
12:14
0:54
1:08:13
11:33
2:19
2:19
2:19
1:16:47